2. p2
7ThingsYouNeedtoDoAboutBigDatain2013
What a year 2012 was for big data! From the White
House to your house, it’s hard to find an organization or
consumer who has less data today than a year ago.
Database options proliferate, and business intelligence
evolves to a new era of organization-wide analytics.
And everything’s mobile.
Organizations that successfully adapt their data
architecture and processes to address the three
attributes of big data — volume, variety and velocity
— are improving operational efficiency, growing
revenues and empowering new business models. With
all the attention organizations are placing on innovating
around data, the rate of change will only increase. So
what should you expect to do to succeed with big data
in 2013?
Simplify
Finding it hard to keep track of all of the new database
vendors and open-source projects? Well, 2013 will even
be more crowded.
Take a strategic approach by extending your relational
and online transaction processing (OLTP) systems to
one or more of the new on-premise, hosted or service-
based database options that best reflect the needs of
your industry and your organization. And pick a real-
time business intelligence platform that supports direct
connections to over 30 database and file formats, with
more every year, to enable you to “pick the right tool for
the job”. Choose the best mix for every project between
connecting live to fast databases and importing data
extracts into an in-memory analytics engine to offset
the performance of slow or overburdened databases.
eBay’s data architecture comprises Teradata, Hadoop
and Tableau. eBay employees can visualize insights
from more than 52 petabytes of data. InformationWeek
writes in December 2012 in “Big data visualization: A big
deal for eBay”: “eBay uses Tableau to visualize search
relevance and quality of the eBay.com site; monitor the
latest customer feedback and meter sentiments on
eBay.com; and achieve operational reporting for the
data warehouse systems. This has helped an analytic
culture flourish within eBay.”
Coexist
Use the strengths of each database platform and
enable them to coexist in your organization’s data
architecture. Cloudera and Teradata jointly published a
useful guide outlining requirements that are best suited
for either a data warehouse or Hadoop.
1
Source: Matthew Aslett, The 451 Group, Updated data-
base landscape graphic, November 2, 2012
Source: Dr. Amr Awadallah and Dan Graham, “Hadoop
and the Data Warehouse: When to Use Which”, co-
published by Cloudera, Inc. and Teradata Corporation.
*HBase.
2
3. p3
7ThingsYouNeedtoDoAboutBigDatain2013
Visualize
According to Forrester Research: “Enterprises find
advanced data visualization platforms to be essential
tools that enable them to monitor business, find
patterns, and take action to avoid threats and snatch
opportunities.” 1
Visual analysis means exploring data visually. A story
unfolds as you navigate from one visual summary into
another. According to Dana Zuber, Vice President of
Strategic Planning at Wells Fargo: “A well-crafted,
thoughtful visualization makes the light bulb go off. You
just don’t get that with a spreadsheet.”
A visual analysis experience has certain characteristics.
It allows you to do two things at any moment:
• Instantly change what data you are looking at – this
is important because different questions require
different data.
• Instantly change the way you are looking at it – this
is important because each view may answer
different questions.
This combination creates the exploratory experience
required for anyone to answer questions quickly.
The cycle of visual analysis is a process of getting data,
structuring it one way, noticing results and asking
follow-on questions. These follow-on questions might
lead to a need to drill down, drill up, filter, bring in new
data, or create another view of your data. And you’ll want
to share and act on the data.
Without interactivity, you and your colleagues are left
with unanswered questions. With interactivity, the data
visualization becomes a natural extension of your
thought process. Furthermore, interactivity allows
colleagues to ask and answer their own questions,
shortening the time it takes to convince colleagues and
moving faster from insights to business results.
Empower
Big data and self-service business intelligence go hand
in hand, according to by Aberdeen Group’s recently
published Go Big or Go Home? Maximizing the Value of
Analytics and Big Data: “Organizations with big data are
over 70% more likely than other organizations to have BI
projects that are driven primarily by the business
community, not by the IT group.” Across a gamut of uses
– from tackling new business problems, developing
entirely new products and services, finding actionable
intelligence in less than an hour, and blending data from
4
See and interact with the live web dashboard on your
desktop, iPad or Android tablet.
1
The Forrester Wave™: Advanced Data Visualization Platforms, Q3 2012, July 17, 2012.
3
4. p4
7ThingsYouNeedtoDoAboutBigDatain2013
disparate sources – “… big data has fired the
imagination of what is possible through the application
of analytics.”
Empower everyone in your organization to use and
benefit from big data. Make sure people can access and
interact with data in dashboards right in their browser, or
on a mobile device. Because the idea that a few select
people will be able to get value from big data is so 2010.
Integrate
“Big data is indeed a big deal” said John Holdren,
Director of the White House Office of Science and
Technology Policy, in a March 2012 forum. Among the
examples cited in a White House fact sheet: the U.S.
Department of Defense is investing US$250 million
annually in a series of big data programs led by the
Defense Advanced Research Projects Agency
(DARPA), including visual analysis of raw and
unstructured data.
Consider how to integrate and blend data from
disparate sources for your organization. Organizations
that can blend different relational, semi-structured and
raw data sources in real time, without expensive
up-front integration costs, will be the ones that get value
from their data. That means that you and your
colleagues don’t need to know the details of how data is
stored to ask and answer questions. Whether your data
is in a spreadsheet, a database, a data warehouse,
open source file systems like Hadoop, or all of those,
you need the flexibility to quickly connect to the data
you need and consolidate it.
The importance of integrating disparate big data is one
area of agreement in the United States for both
Democrats and Republicans. “The Obama campaign
found a way to integrate social media, technology, email
databases, fundraising databases and consumer
market data,” said GOP digital strategist Vincent Harris,
who did digital work for Newt Gingrich and Rick Perry in
2012. “That does not exist on the Republican side to
that degree”, to the detriment of Mitt Romney’s
campaign, quoted by Politico, “GOP seeks to up its
online game”, December 8, 2012. For more on how the
Obama campaign used big data, see BusinessWeek’s
November 29, 2012 article “The Science Behind Those
Obama Campaign Emails”.
Govern
More than 80 countries have data privacy laws. The
European Union defines seven “safe harbor privacy
principles” for the protection of E.U. citizens’ personal
data. In Singapore, the personal data protection law
takes effect January 2013. In the U.S., Sarbanes-Oxley
affects all publicly listed companies, and HIPAA (Health
Insurance Portability and Accountability Act) sets
national standards in healthcare.
The right balance between control and experimentation
will vary depending on your organization and industry.
It’s never an either / or – you’ll need both – but within
the constraints of your industry and your corporate
policies.
If your organization has already implemented
successful master data management (MDM),
congratulations. But if you’re like most organizations,
Source: Tableau Software, Mobile Business Intelligence
5
6
5. p5
7ThingsYouNeedtoDoAboutBigDatain2013
you’ve found agreement on definitions and business
rules for master data management to be a slow, painful
process for IT and business leads. A group of MIT-
Sloan Management Review authors provide three tips in
“Finding Value in the Information Explosion”: identify
your “sacred data”; define the workflows that will use
unstructured data; and use data to redefine business
processes following a highly iterative process.
The contradiction of metadata is that you need to both
centralize data sources and apply metadata, yet still be
able to extend it by adding your own calculations;
creating new groups; organizing data into hierarchies;
and modifying aliases. Organizations that master big
data will let IT set up the data architecture, security and
access controls, while giving business people the ability
to serve themselves reports and dashboards.
Evangelize
With the backing of one or more executive sponsors,
evangelists like yourself get the ball rolling and instill a
virtuous cycle – the more departments in your
organization that realize actionable benefits, the more
pervasive analytics becomes across your organization.
Fast, easy-to-use visual analytics is the key that opens
the door to organization-wide analytics adoption and
collaboration.
For Blastrac Manufacturing CIO Dan Murray: “Before,
we were spending four to six man-hours a week
producing basic reports. Now it takes us ten minutes a
week. And the quality of the report is much better, much
more visible and interactive.” See case studies for your
industry.
So there you have it, how to succeed with big
data in seven points.
7